Hard to believe how much progress we’ve had since the OG Pepperoni Hug Spot pizza chain TV ad (April 2023) and the Will Smith eating spaghetti (March 2023) - names and dates according to ChatGPT 😊
Incredible! If it weren't for the odd tiny artifact mistake here and there, I would be 100% convinced that this was real video. This is going to change entertainment forever. As an actor, this terrifies me....
"Hi, how long did you wait for access? I've been waiting for about two weeks, since the very beginning. Could you please tell me when you applied for early access and when you received it?
@@kratos5955 Nah, how can a few illuminated dots on a modern display be harmful or dangerous. It's just an illusion after all🤷♂TBH CRTs were much more harmful due to the radiation.
I know that such statements are always dangerous if you haven't tried a model yourself, but google's veo 2 is what open ai promised with Sora. And when it comes to video models, google's hypersensitive moderation strategy might actually be a good thing for once.
I used something along the lines of: "Vlog like shot, a woman with blond and pink hair speaks to the camera, the woman is walking through a european style street during the day."
@@jerrod_lew Ok thanks. How does this keep up with new trends/fashions/current events? Google constantly grab from the internet? Or they do it more officially and pay people to provide content?
This just looks like real people tbh. I kinda feel like it's just using memory of real people rather than create it's own tbh. How much customization do you have when prompting how people look? I have to actually freeze frame it to see the inconsistencies.
do you work for google? how come you are the only one who has access? Also do you have any idea how much it will cost if it is going to be 200$ month like sora this is very disappointing!
@@Lofi-NPC-girl I did join but they did not accept my request yet, and I know it is free for now I m sure that's going to change when it is going public
I wish I worked for Google! And a lot of people have access from the wait list I think, not sure when they will roll it out more. I haven't seen anything about the pricing yet.
I’ve seen a bunch of people that have access, but still is pretty limited. Seems bigger influencers more likely to get access and then is a bit more random after that.
Can I upload a picture of him and he can move it?? Can he output the movement and scenes accurately as written in the text??? Can he move a picture with the required movements with the possibility of uploading an audio file
Thankfully you can still tell by pausing the video every now and then to see the inconsistencies. But it's scary just last year we could easily tell by the random morphing.
@@jerrod_lew For example, it would be interesting to see if this tool can include story in these clips. I mean create a clip that has a clearly different beginning and ending. Most tools seem to struggle with this big time atm
@@ElfTaleFilms Well, the tool only generates short clips (8 seconds), and consistency is still a challenge. Howvever, with Veo there is some coherence. Check out this video that involves consistent scene cuts: th-cam.com/video/Q2Xk7vXSnUg/w-d-xo.html&ab_channel=JerrodLew
@@jerrod_lew I don't mean consistency. 8 second clip is enough to show change but these tools just are terrible at it! You can achieve consistency between clips already with certain tricks, but you can't have beginning and ending be really that different, making it not so useful for story tellers.
@@jerrod_lew Are you kidding me? This is incredibly disturbing to watch and last thing I want is that the internet gets filled with even more fake influencer videos than we have now.
@@astralyd Well it's a different situation for everyone, and using AI is optional, and seeing AI content is also optional. This kind of technology can help people express themselves creativity in new ways, it can reduce costs of content creation/marketing and advertising. It can also give people a creative voice without technical barriers. AI has the potential to give people more choices and opportunities. And the best part is you can completely avoid it if you want to.
@@jerrod_lew How is this optional? I've already had a ton of this stuff recommended to me, and I have a feeling it's only going to get worse from here. Basically, you're stealing other people's videos for money, that's how you reduce the cost, not to mention all the other problems it causes. it's just laziness, just learn to produce authentic content if you're worth anything, this trend should stop as soon as possible. GPT Chat is good for solving problems, but this on the other hand is horrible, I hope I'm not risking anything by saying this.
Hard to believe how much progress we’ve had since the OG Pepperoni Hug Spot pizza chain TV ad (April 2023) and the Will Smith eating spaghetti (March 2023) - names and dates according to ChatGPT 😊
It's really crazy how far we have come so fast.
Exponentials
Still we have little control and end up in random short clips with scenes that never to be seen again
@ I agree
What did I just watch, this is insane!!! Had to double check if you put an actual video of a real person. Thanks for the video❤ great idea
Thanks for watching!
For some reason the image of Cell from DBZ when he says "Incredible..." springs to mind.
Incredible! If it weren't for the odd tiny artifact mistake here and there, I would be 100% convinced that this was real video.
This is going to change entertainment forever.
As an actor, this terrifies me....
"Hi, how long did you wait for access? I've been waiting for about two weeks, since the very beginning. Could you please tell me when you applied for early access and when you received it?
The fidelity is unreal
This is unbelievably scary
Correction: this is unbelievably amazing!
@TheGalacticIndian I do not need a person who does not know that ai video generators can also be harmful and dangerous to come and correct me.
@@kratos5955 Nah, how can a few illuminated dots on a modern display be harmful or dangerous. It's just an illusion after all🤷♂TBH CRTs were much more harmful due to the radiation.
Holy shit so this is why ytb asked me to allow them taking references from the videos on my channel.
This really incredible
I know that such statements are always dangerous if you haven't tried a model yourself, but google's veo 2 is what open ai promised with Sora. And when it comes to video models, google's hypersensitive moderation strategy might actually be a good thing for once.
it can do cuts, how
This is what shocks me, it seems so coherent to the scene.
Hi Jerrod, Can you provide the prompt text for this video?
I used something along the lines of:
"Vlog like shot, a woman with blond and pink hair speaks to the camera, the woman is walking through a european style street during the day."
@@jerrod_lew Great,Thank you
I guess the prompts are leaning on what's in the dataset though, right? Would be interested to see the less successful results too for comparison.
I'm posting a good range of results, they are honestly super strong in general
@@jerrod_lew Ok thanks. How does this keep up with new trends/fashions/current events? Google constantly grab from the internet? Or they do it more officially and pay people to provide content?
This just looks like real people tbh. I kinda feel like it's just using memory of real people rather than create it's own tbh. How much customization do you have when prompting how people look?
I have to actually freeze frame it to see the inconsistencies.
You can get pretty diverse with the prompts, hair color, eye color, ethnicity, origin, clothing, environments etc.
How to access veo2 not available in my country
You can try to sign up to the waiting list by using a VPN for USA.
I would like to see a similar pace of development with AI self-driving cars.
do you work for google? how come you are the only one who has access? Also do you have any idea how much it will cost if it is going to be 200$ month like sora this is very disappointing!
Its free
Jion the witing list
@@Lofi-NPC-girl I did join but they did not accept my request yet, and I know it is free for now I m sure that's going to change when it is going public
I wish I worked for Google! And a lot of people have access from the wait list I think, not sure when they will roll it out more. I haven't seen anything about the pricing yet.
I’ve seen a bunch of people that have access, but still is pretty limited. Seems bigger influencers more likely to get access and then is a bit more random after that.
Can I upload a picture of him and he can move it?? Can he output the movement and scenes accurately as written in the text??? Can he move a picture with the required movements with the possibility of uploading an audio file
There is image to video, but you have to use the Imagen 3 engine to create the imagine first. No audio uploads yet.
Awesome big Google
It's scary good. Everything will change.
We are just getting started.
@jerrod_lew Right? I don't mean "scary" as in I would be opposed to it. It just moves at vertigo inducing speed.
The first people who do Ai vlogs right, will make a lot of money.
Agreed!
"This cant be... unreal" wow
Thankfully you can still tell by pausing the video every now and then to see the inconsistencies. But it's scary just last year we could easily tell by the random morphing.
AI has been able to produce 100% realistic images for several months already. The real question is, can these tools handle real use case scenarios?
What would be a real use case scenario for you?
@@jerrod_lew For example, it would be interesting to see if this tool can include story in these clips. I mean create a clip that has a clearly different beginning and ending. Most tools seem to struggle with this big time atm
@@ElfTaleFilms Well, the tool only generates short clips (8 seconds), and consistency is still a challenge. Howvever, with Veo there is some coherence.
Check out this video that involves consistent scene cuts: th-cam.com/video/Q2Xk7vXSnUg/w-d-xo.html&ab_channel=JerrodLew
@@jerrod_lew I don't mean consistency. 8 second clip is enough to show change but these tools just are terrible at it! You can achieve consistency between clips already with certain tricks, but you can't have beginning and ending be really that different, making it not so useful for story tellers.
AGI will be able to feed stupidity, great😂
how did this guy get into early access,,,dafuq
I can just recommend being on the waitlist and join their Discord.
We're cooked
I mean wow crazy do any of these people actually exist?!
Nothing in this video is real, all raw footage from Veo 2!
what's the point of this
To give people more ways to create content.
@@jerrod_lew Are you kidding me? This is incredibly disturbing to watch and last thing I want is that the internet gets filled with even more fake influencer videos than we have now.
@@astralyd Well it's a different situation for everyone, and using AI is optional, and seeing AI content is also optional.
This kind of technology can help people express themselves creativity in new ways, it can reduce costs of content creation/marketing and advertising.
It can also give people a creative voice without technical barriers.
AI has the potential to give people more choices and opportunities. And the best part is you can completely avoid it if you want to.
@@jerrod_lew How is this optional? I've already had a ton of this stuff recommended to me, and I have a feeling it's only going to get worse from here. Basically, you're stealing other people's videos for money, that's how you reduce the cost, not to mention all the other problems it causes. it's just laziness, just learn to produce authentic content if you're worth anything, this trend should stop as soon as possible. GPT Chat is good for solving problems, but this on the other hand is horrible, I hope I'm not risking anything by saying this.
@jerrod_lew My reply has just been deleted. It's crazy what's happening. It's so shameless. We are doomed
we're doomed
Opportunities are always there.
Lmaoo
The mouths don't quite close enough to look normal.
We won't need jobs or entertainers. nirvana.
Can it generate game footage?
I would also like to see this.
Currently testing this!